US8773055B2 - One-phase electronically commutated motor - Google Patents

Abstract

An electronically commutated one-phase motor (20) has a stator having at least one winding strand (30, 32) and a permanent-magnet rotor (22). The rotor, as it rotates, induces a voltage (u_ind) in the at least one winding strand (30, 32). The motor has an electronic calculation device (26), preferably a microcontroller μC, which is configured to execute, during operation, the steps of a) sensing the value of the instantaneous operating voltage (Ub); (b) using the operating voltage value (Ub) and optionally further parameters, adjusting a time duration (T_ON) of a switch-on current pulse (i30) for the motor, in order to apply a consistent amount of electrical energy to the windings during start-up attempts, thereby maximizing the probability of successful start-up, regardless of possible fluctuations in motor operating voltage and related operating parameters. The switch-on current pulse duration (T_ON) can be adjusted longer or shorter, as a function of operating experience.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

This application is a continuation-in-part of our U.S. application Ser. No. 12/830,492 filed 6 Jul. 2010, now U.S. Pat. No. 8,294,398, and also claims priority from ourGerman application DE 10 2010 004 361.1 filed 12 Jan. 2010.

FIELD OF THE INVENTION

The present invention relates generally to an electronically commutated motor (ECM) and, more particularly, to an ECM with measures to make its start-up more reliable.

BACKGROUND

One-phase electronically commutated motors are very inexpensive, and are often used for specific driving tasks, e.g. for fans or centrifugal pumps. They are usually controlled by means of a Hall sensor which magnetically detects the instantaneous rotational position of the rotor. Commutation without a sensor, referred to using the term “sensorless,” is, however, desirable, since better efficiency is obtained as a result.

The terminology of such motors is somewhat confusing. For accurate definition of an ECM, firstly the number of stator current pulses per rotor rotation of 360° el. is indicated, e.g. single-pulse, two-pulse, three-pulse, etc.; also the number of winding strands in the stator is indicated, e.g. single-strand, two-strand, three-strand, etc.

An ECM can therefore e.g. be described as single-strand and two-pulse, or two-strand and two-pulse. The expression “collectorless motor” is also used instead of “ECM”. Because there is no difference between the single-strand and two-strand motors in terms of physical operation, and because simplified terminology is always desirable for practical use, such motors are generally referred to as “one-phase” ECMs, even though they can have either only a single strand or, alternatively, two strands.

Because the rotor in such motors has rotational positions at which the motor cannot generate any electromagnetic torque, an auxiliary torque is used that is effective at those zero positions. This can be a magnetically generated auxiliary torque, which is referred to as reluctance torque. Alternatively, this auxiliary torque could be generated mechanically, for example by means of a spring that is tensioned in certain rotational positions and delivers its stored energy at the zero positions. The result is that the rotor, at a standstill, is rotated sufficiently that at startup it is not in a rotational position in which the motor cannot generate an electromagnetic torque, since otherwise the motor would not be able to start. This starting position is also referred to as a “cogging” position.

When such motors are currentless, normally the rotor is at a standstill and is in a so-called cogging position, into which it is pulled by the aforesaid auxiliary torque. When current is applied to the motor with the rotor in this position, the rotor will move; it is, however, only possible to guess how strongly it will move.

SUMMARY OF THE INVENTION

It is therefore an object of the invention to make available a novel electronically commutated motor (ECM) with a reliable minimum rotor movement during start-up.

According to the invention, this object is achieved by sensing the instantaneous motor operating voltage, and using this value to adjust the time duration of a switch-on current pulse applied to the motor. Sensing of the operating voltage creates the possibility of correctly metering energy delivery at startup. This is because energy is delivered at startup as a so-called current-flow block and, by means of the invention, this block can be metered so that, regardless of the instantaneous operating voltage, approximately the same quantity of energy is delivered at startup. Subsequent thereto, a check is made as to whether that quantity of energy is to small or too large, and corresponding corrective measures are taken, as appropriate, to apply a consistent quantity of energy.

BRIEF FIGURE DESCRIPTION

Further details and advantageous refinements of the invention are evident from the exemplifying embodiments, in no way to be understood as a limitation of the invention, that are described below and depicted in the drawings.

FIG. 1 is an overview diagram that schematically shows various situations that can occur during operation of a sensorless one-phase motor and must therefore be taken into account in its software, in order to ensure reliable starting;

FIG. 2 is a schematic depiction to explain an ECM that operates with a reluctance torque;

FIG. 3 is a circuit diagram of an embodiment of a one-phase motor that is configured to take into account situations ofFIG. 1, the motor being illustrated as a two-strand motor;

FIG. 4 is a depiction to explainFIG. 3; and

FIG. 5 shows a routine to optimize a time period Tv whose length is important for optimum commutation.

DETAILED DESCRIPTION

The diagram ofFIG. 1 schematically shows problems that must be taken into account, when developing a “sensorless one-phase ECM.”

After being switched on, the motor is insituation10 ofFIG. 1, i.e. it has either a rotation speed n=0 or (if externally driven) a rotation speed n≠0, and its rotational position “pos” is unknown (to the motor's electronic control circuit), since a rotational position sensor is not present.

According tosituation12, this can mean that the rotation speed is zero, and the rotor is in one of its cogging positions that is defined by the cogging torque.

The motor can, however, also be insituation14, meaning that an external driving force is acting on it. In the case of a fan, for example, a wind gust or a storm can be driving the motor, so that, although it is receiving little or no motor current, its rotation speed n is nevertheless not equal to zero, since the rotor (in the case of a fan) can be driven like a windmill by a storm.

Under these conditions, however, the motor can also run in both rotation directions (seepositions16 and18 ofFIG. 1) while the normal motor current is flowing. If the motor is rotating in the wrong rotation direction, corresponding countermeasures are then necessary. The “wrong” rotation direction thus means that reversing must occur after startup.

Step14 measures whether an induced voltage u_indis present, i.e. whether the magnitude of u_indis greater than zero. This can also be the case, for example, when a fan is being passively driven by wind. In addition, a measurement is made as to whether the magnitude of rotation speed n is greater than zero.

If the response to both queries is NO, the program goes tostep12, which indicates that the rotation speed has a value of zero, and also that the rotational position of the rotor is defined by the so-called cogging torque, i.e. the rotor has “snapped” into one of its cogging positions.

If the responses instep14 are YES, the rotor is either rotating in its preferred direction PRDIR (step16) or rotating oppositely to its preferred direction PRDIR (step18). The rotation direction cannot, however, be immediately deduced from the available data.

The motor can, however, also rotate in either of the two rotation directions as a result of external influences; the normal motor current is flowing, but the motor can rotate in the wrong direction. The “wrong” rotation direction means that it must be reversed after starting.

FIG. 2 shows the circuitry of an Electronically Commutated Motor (ECM)20 that operates in sensorless fashion.Motor20 has a permanent-magnet rotor22 (indicated merely schematically) that is depicted with four poles, but can also have four, six, eight, etc. poles.Rotor22 can be an internal rotor, external rotor, the rotor of a motor having a flat or conical air gap, etc.

Motor20 has a microcontroller (μC)26, preferably a PIC12F629 from Microchip Technology, Inc., Chandler, Ariz., 85224, USA. It can include aROM74,RAM76, and timer80, as shown inFIG. 2. The power supply system ofμC26 is, as usual, not depicted.Motor20 has twostator winding strands30,32 that are usually magnetically coupled via the magnetic circuit of the motor, as indicated bysymbol34′. Placed in series withfirst winding strand30 is a first semiconductor switch, here e.g. an n-channel MOSFET (Metal Oxide Semiconductor Field Effect Transistor)34, which has arecovery diode38 connected antiparallel to it and which is controlled byμC26 via acontrol line36. Together withsemiconductor switch34 anddiode38,strand30 forms afirst series circuit40 that can optionally contain further elements.

Located in series withsecond strand32 is a secondcontrollable semiconductor switch44 that is controlled byμC26 via acontrol line46. This switch can likewise be an n-channel MOSFET44 that has arecovery diode48 connected antiparallel to it. Together withsecond semiconductor switch44,second strand32 forms asecond series circuit50 that may contain further elements.

AsFIG. 2 shows, the twoseries circuits40,50 are connected in parallel, to form aparallel circuit53 whosebottom node55 is connected to ground56. The upper ends ofstrands30,32 are also connected to aDC link circuit58. This means that whensemiconductor switch34 is conductive, a current i30 flows fromlink circuit58 throughfirst strand30, and whensemiconductor switch44 is conductive, a current i32 flows throughstrand32. This statement must, however, be modified for the time intervals just before a commutation, as will be explained below.

Link circuit58 is connected via a third semiconductor switch60 (here a p-channel MOSFET) to amotor terminal62 to which a positive voltage Ub, e.g. 12, 24, 48, 60 V, etc. is applied towardground56 during operation. ADC source63 of any kind is depicted symbolically. A power supply, for example, often serves as a DC source.

Adiode61 can be located antiparallel tothird semiconductor switch60.Third semiconductor switch60 is controlled byμC26 via acontrol line64.

A potential from drain D ofsemiconductor switch34 is delivered to acomparator input65 ofμC26 through asensor line66 and aresistor67.Input65 is connected via aZener diode69 toground56, in order to protect said input from overvoltage.

A potential from drain D ofsecond semiconductor switch44 is likewise delivered to acomparator input71 ofμC26 through asensor line68 and aresistor73.Input71 is connected via aZener diode69′ to ground56, in order to protectinput71 from overvoltage.

In addition, a voltage divider, made up of tworesistors75,76′, whose connectingnode77 is connected to input A/D of an analog-to-digital converter inμC26, is connected between drain D offirst semiconductor switch34 andground56.

Measuring Ub

This measurement is made viavoltage divider75,76′. The latter of the two resistors is dimensioned so that the internal reference voltage (in thiscase 5 Volts) of the A/D converter inμC26 cannot be exceeded. This precludes measurement errors. Alternatively, this voltage divider can also be placed between source S ofthird semiconductor switch60 andground56.

Voltage divider75,76′ also, additionally, has another function: depending on the amplitude of the voltages that are induced instrands30,32, said voltages are limited byprotective diodes69,69′. It is important for rotation direction detection, however, to sense the shape of the induced voltages atinputs65 and71, respectively, which would be prevented by voltage limiting. In this instance, the induced voltage is therefore sensed byvoltage divider75,76′ and input A/D of μC27, with the result that the shape of the induced voltage can also be detected.

The signals at drains D offirst semiconductor switch34 and ofsecond semiconductor switch44 are sensed atcomparators65,71 inμC26.

Manner of Operation ofFIG. 2

Reference is made for this purpose toFIG. 3. Shortly before time instant t0 inFIG. 3, all threesemiconductor switches34,44,60 inFIG. 2 are blocked, andmotor20 consequently receives no energy fromterminal62, i.e. energy delivery fromDC source63 is blocked.

At time to,transistors34,60 are switched on byμC26 so that a current i30 flows from terminal62 throughtransistor60,link circuit58,strand30, andtransistor34 toground65.FIG. 3a) shows the shape of current i30, which of course depends on the value of the motor rotation speed and on other factors.

Commutation time instant t0 is followed by a commutation time instant t4, at whichtransistor34 is switched off andtransistor44 is switched on, so that current i30 is shut off and current i32 (through strand32) is switched on.

Located in a time interval Tv before t4 is a time instant t2 at whichtransistor60 becomes blocked, so that energy delivery fromterminal62 is interrupted, i.e. no energy is delivered fromDC source63 tomotor20 during this time period Tv.

A specific current I flows instrand30 shortly before time t2, so that a specific energy E is stored instrand30 in accordance with the formula
E=0.5*L*i²  (1)
where
E=energy stored in the magnetic field of the relevant strand
L=inductance of that strand
I=current at time instant t2.

This stored energy now causes a loop current I* to flow throughstrand30 becausetransistor34 is still conductive. This loop current I* flows from the lower terminal ofstrand30 throughtransistor34,node55,recovery diode48, and the twostrands32 &30 so that, as before, it generates a driving torque onrotor22, with the result that loop current I* rapidly drops and, at time instant t3 ofFIG. 3 reaches a value of zero.Transistor34 can therefore be blocked in a wattless manner as of time instant t3, since loop current I* has become zero.

Measuring Operating Voltage Ub

It is important for startup purposes to know the operating voltage Ub of the motor. While most ECMs have a fixed operating voltage, they can also be operated in an extended range of operating voltages. ECMs for a voltage of 48 V should therefore be able to be operated within a voltage range extending approximately from 36 V to approximately 72 V. The result of these voltage differences is to produce, for the same current-flow duration, very different rotor accelerations for the first current-flow block: the rotor is more strongly accelerated at higher operating voltages, and it may happen that the change in induced voltage is therefore not detected, so that commutation cannot take place. In order to ensure detection of the induced voltage after the first current-flow block, the first current-flow blocks must be adapted with respect to the operating voltage. When the motor is at a standstill, the operating voltage can be identified very easily by way ofvoltage divider75,76′ and the A/D converter inmicrocontroller26 which digitizes the analog operating voltage signal from the voltage divider. This can be done by switching onsemiconductor switch60 and switching off the twopower stage transistors34 and44. In this case, the operating voltage is measured directly at drain terminal D ofpower stage transistor34.

If the motor is additionally being driven from outside (referred to as an “external driving force”), an induced voltage additionally occurs at the measuring points of the winding. This voltage is overlaid on the operating voltage, and the latter therefore cannot reliably be detected. In this case, the operating voltage can be measured at a point upstream fromtransistor60. This variant, however, requires additional circuit complexity. One solution is to proceed in such a way that when an external driving force is identified, the induced voltage and its zero crossings are observed. This is possible if all three semiconductor switches are left nonconductive, with the result that no operating voltage is present at windingstrands30,32, and the change in induced voltage can be observed. When a zero crossing of the induced voltage then occurs,transistor60 is switched on and the operating voltage is then present at the winding strands. At that moment, the induced voltage has no influence, and only the operating voltage is measured. Depending upon the rotation speed, this method must function very quickly, since before and after the zero crossing the induced voltage has a steep edge slope that might otherwise cause measurement errors.

In addition to the lack of information about rotor position, there are other important factors that require attention in the context of a sensorless startup. Different winding resistance values, and hence different winding currents and different startup torques, occur, depending upon the operating voltage, winding design, and winding temperature. The startup torque is opposed by frictional torques that change with temperature and with the age of the motor. Attention must also be paid to the differences in axial moment of inertia among different rotors. Different angular accelerations are also produced depending on the rotation direction.

If the angular acceleration achieved by means of the first current-flow block is sufficient for evaluation of the voltage induced by the rotor, this makes it possible to ascertain the rotation direction in which the rotor was accelerated. This means that the first current-flow block that is selected must not be too long, so that after current flow and after the subsequent current loops, the induced voltage can also be measured.

If an induced voltage cannot be measured after the first current-flow block (during time period T_onofFIG. 3), there may be various reasons for this. The motor may have been blocked by an external influence; or the first current-flow block was set too short, and as a consequence of increased bearing friction, aging, or a high winding resistance, the electrical torque generated would then not be enough to accelerate the rotor sufficiently. In this case, an induced voltage cannot be identified. Provisions must be made for all these different instances.

FIG. 4 shows a starting routine for the normal case, in whichrotor22 is in a predetermined rotational position from which it is to be started.

Starting occurs at S250. S252 checks whether the induced voltage Uind differs from zero, i.e. checks whetherrotor22 is rotating. If YES, the program goes to a special routine254 for startup, and then (at S256) transitions into a standard commutation in the desired rotation direction. One such commutation is described below with reference toFIG. 5.

If the response at S252 is No, the program goes to step S258, wheretransistor60 is switched on andtransistors34 and44 are switched off, in order to measure the operating voltage Ub at input A/D ofμC26.

In the next step S260, a factor x is derived (for example from stored tables) from Ub and optionally from other factors or parameters, e.g. the instantaneous temperature, and in S262 the operating voltage Ub is multiplied by this factor in order to calculate the duration T_onof the switch-on current pulse that is calculated on a predictive basis for startup ofmotor20.

The twotransistors60 and34 are then switched on at S264, with the result that current i30 throughstrand30 is switched on androtor22 is accelerated.

After time T_oncalculated in step S262 has elapsed, instep S266 transistor60 is switched off, thereby interrupting energy delivery fromcurrent source63. But becausetransistor34 is still conductive, the magnetic energy stored in windingstrand30 causes a loop current I* to flow fromnode54 throughtransistor34,recovery diode48, and the two windingstrands32 and30 back tonode55, and this loop current I* drivesrotor22 and thereby rapidly drops to zero.

As long as loop current I* is flowing, drains D of the twotransistors34 and44 are at ground potential; but when I* has become equal to zero, an induced voltage u_indindicating the rotation ofrotor22 is obtained at drain D oftransistor34. This voltage is sensed in step S268.

If it is not possible to sense any such induced voltage, the program goes to step S270, where time span T_onis extended by an amount equal to an “Offset” value; the program then goes to step S264 in order to repeat the startup attempt at an increased energy.

If the response in S268 is YES, S272 checks whether the induced voltage at the drain oftransistor34 can be measured. If NO, time span T_onis too long, and in S274 it is therefore shortened by an Offset correction time, in order to weaken the startup current pulse.

The program then goes to step S264 in order to repeat the startup operation at reduced energy. If, however, the response in S272 is YES, i.e. if the induced voltage does occur at the drain oftransistor34, this means that the loop current has dropped to zero at the correct time, and the program goes to step S276 whereECM20 is commutated normally. In this case,motor20 is running normally, andmotor20 usually starts without difficulty.

Optimizing Commutation Time t4

Optimized commutation is important for optimum and low-loss operation ofmotor20, since the motor then runs quietly with good efficiency.

Commutation optimization is of course particularly difficult with a sensorless motor because a rotor position sensor is not present, so that optimization requires working with other variables that can be measured.

FIG. 3 shows at a) the currents i30, i32 in the twostrands30 and32 ofmotor20. The potential p52 atnode52 ofFIG. 2, i.e. at drain D ofFET44, is depicted at b). Because the arrangement is symmetrical, the potential p54 atnode54 has the same profile but offset 180°, and is therefore not depicted inFIG. 3.

As long asFET44 is conductive, its drain D is connected to ground56, so that a voltage induced by the permanent-magnet rotor22 instrand32 cannot be measured atnode52.

As soon as current I* has dropped to zero, however, this induced voltage (labeled inFIG. 3b) as68) can be measured atnode52, so that the occurrence ofvoltage68 means that loop current I* has dropped to zero; this is the case at time t3, and means that as of that time wattless commutation can take place.

Time span Tv, between time t2 at whichFET60 becomes blocked and current i30 is thereby switched off, and time t4 at whichFETs44 and60 are switched on, so that a current i32 which flows throughstrand32, therefore has an optimum value when time span Tp between times t3 and t4 becomes as short as possible, since Tv then also has a minimal value.

On the other hand, of course, Tv must not become too short, since then the switching on of current i32 (time t4) would fall in a time period Ti (between t2 and t3) in which loop current I* is still flowing, so that wattless commutation would not be possible.

In this case, time span Tv must therefore be extended.

The operations depicted in the flow charts ofFIGS. 4 and 5 serve this purpose.

Time Tv (FIG. 3), which is set at the startup ofmotor20 to a default value, and at the beginning of which (at time t2) the “prelude” to each commutation begins, can be optimized by means ofμC26. The corresponding routine is depicted inFIG. 5.

This routine begins at step S88 and is preferably called at each commutation. In S88, Tv is set to a default value after switching on. The optimization of Tv begins in S90. S92 checks whether the end (t3) of current looping was detected before commutation time t4. If so, Tv is reduced in S94 by an decrement ΔTv1. If not, then in S96 Tv is increased by an increment ΔTv2 that is greater than decrement ΔTv1 in step S94. Optimization ends at step S98.

The result is that an optimum value for Tv is automatically established within a few revolutions, even if the motor rotation speed has changed as a result of external influences, e.g. an air current.

Problems at Higher-Order Transistor60

At startup or in the event of a change in the load onmotor20, it may happen that the higher-order transistor60 becomes blocked too late, and a loop current is therefore still flowing throughstrands30,32 at the commutation time. A currentless commutation is not possible in such a case, and protective measures must be taken to prevent this.

One possibility here is to use a link circuit capacitor, which is arranged betweenlink circuit58 andground56 and which absorbs the residual magnetic energy of the winding strand that is to be switched off and thereby limits the voltage atlink circuit58.

It is also possible to insert a Zener diode betweenlink circuit58 andground56, in order to limit the voltage atlink circuit58.

The drain voltages ofFETs34 and44 can also be limited, using respective Zener diodes that are arranged between the pertinent drain D andground56.

Another, and possibly additional, action is to limit the drain voltages ofFETs34,44 by slow switching. This can be achieved using a series circuit of a capacitor and a resistor that is connected between drain D and gate G of the relevant FET.

The drain voltages ofFETs34,44 can also be limited by slow switching of the relevant FET. This can be done using a series circuit of a Zener diode124 and a resistor126. In this case, a series circuit of this kind is inserted between D and G of the relevant transistor.

Current limiting can additionally be provided formotor20. This is not depicted inFIG. 2, so as not to make the depiction difficult to understand as a result of a proliferation of elements. Current limiting is preferably achieved by blocking higher-order transistor60 in the event of an overcurrent, in order to interrupt energy delivery fromDC source63 tomotor20. This results in a respective loop current I* as already described, and this loop current generates a torque, thus yielding current limiting with highly efficient motor operation.

Many variations and modifications are, of course, possible within the scope of the inventive concept.

Claims (12)

What is claimed is:

1. An electronically commutated one-phase motor (20) comprising:

a stator having a permanent-magnet rotor (22) and two winding strands (30,32), each of which is connected in series with a respective associated semiconductor switch (34,44), thereby forming first and second series circuits (40,50) having, upstream from them, a third semiconductor switch (60) which, when blocked, interrupts current delivery from a DC power source (63) to the motor (20),

at least one of the two winding strands having a first end and a second end, the first end being connected to a supply lead (58) for supplying an operating voltage from said DC power source (63), the second end being connected to an associated semiconductor switch (34,44),

wherein the at least one permanent-magnet rotor (22) induces, as it rotates, a voltage (u_ind) in said at least one of the two winding strands (30,32);

an electronic calculation device (26) which is configured to execute, during operation, the steps of:

a) measuring an analog instantaneous operating voltage (Ub) of the motor and deriving therefrom a digital voltage value characterizing said instantaneous operating voltage;

b) specifying a time duration (T_ON) of a switch-on current pulse (i30) applied to the motor, based upon at least one motor operating parameter, including as a parameter said digital voltage value;

c) switching space on said third semiconductor switch (60) for said time duration (T_ON) to apply said switch-on current pulse (i30) to a predetermined winding strand (30) of the motor, to cause a current flow in said predetermined winding strand (30);

d) after said time duration (T_ON) has elapsed, interrupting energy delivery to the motor so that, as a result of magnetic energy stored in the motor (20), a current (I*) is briefly generated, that loops in the motor and flows through the predetermined winding strand (30), which current is referred to hereinafter as a loop current;

e) monitoring as to whether or not an induced voltage occurs after the end of the loop current (I*);

f) if no induced voltage is detected, extending the time duration (T_ON) of the switch-on current pulse (i30), and repeating, during that extended time duration, a starting operation;

g) if an induced voltage does occur after the end of the loop current (I*), making a check as to whether said voltage is occurring at the predetermined winding strand (30);

h) if No, shortening the time duration (T_ON) of the switch-on current pulse, and using that shortened time duration, repeating a starting operation;

I) if Yes, executing a transition to normal commutation of the motor (20).

2. The motor according toclaim 1, wherein said step of specifying the duration of the switch-on current pulse comprises specifying said duration, based upon both said instantaneous operating voltage and motor temperature.

3. The motor according toclaim 1, further comprising

a respective recovery diode (38,48,61), associated with at least one of the semiconductor switches (34,44,60).

4. The motor according toclaim 1, wherein

the semiconductor switches (34,44,60) are implemented as field effect transistors.

5. The motor according toclaim 1, wherein

said electronic calculation device (26) is a microcontroller (26) configured to control the semiconductor switches (34,44,60).

6. The motor according toclaim 5, wherein

during operation, signals (p52, p54), outputted from the semiconductor switches (34,44) associated with the respective winding strands (30,32), are applied (65,71) to the microcontroller (26) for evaluation.

7. The motor according toclaim 6, wherein at least one of the signals from the semiconductor switches (34,44) is applied (65,71) to an A/D converter provided in the microcontroller (26), for the purpose of digitizing the value of said analog instantaneous operating voltage (Ub).

8. The motor according toclaim 7, wherein during delivery of the signal (p52, p54) serving for measuring of the operating voltage (Ub), current flow, to the at least one winding strand (30,32), is interrupted.

9. The motor according toclaim 1, wherein the electronic calculation device (26) is configured to measure an elapsed-time interval (Tp) between the end of said loop current (I*) and a commutation instant (t4) subsequent thereto.

10. The motor according toclaim 9, wherein the electronic calculation device (26) is configured, upon occurrence of said elapsed time interval (Tp), to decrease of said elapsed time interval, in steps.

11. The motor according toclaim 9, wherein

the electronic calculation device (26) is configured, in the absence of a time interval (Tp) of this kind, to modify (S96) at least one commutation parameter in a direction such that a time interval (Tp) of this kind will occur.

12. The motor according toclaim 10, wherein

the electronic calculation device (26) is configured, in the absence of a time interval (Tp) of this kind, to modify (S96) at least one commutation parameter in a direction such that a time interval (Tp) of this kind will occur.

Images